1. Field of the Disclosure
The present disclosure relates to the field of switched-network services.
2. Background
Switched-networks, such as digital subscriber line (DSL) networks, are used to provide services, such as Internet access, Internet Protocol Television (IPTV), Video-on-Demand (VoD), Voice over Internet (VoIP), etc. A typical DSL network includes various servers and routers that form a backbone for the DSL network. These servers receive, among other things, DSL content (such as video, voice and data) from various sources and transmit the DSL content to central offices (COs) located in different geographical areas. Switches, such as Digital Subscriber Line Access Multiplexers (DSLAMs), distribute the DSL content to customer equipment, and establish two-way communication between the network and the customer equipment. A DSLAM typically contains multiple (a few hundred) ports, each providing a dedicated connection to a subscriber or customer interface equipment, such as a modem (also referred to herein as the customer gateway).
In a switched IPTV network, each channel of IPTV service is associated with an IP multicast group and delivered to a customer as an IP multicast stream. A customer typically selects to watch a channel by using a set-top-box (STB) that sends a “JOIN” message to the network, requesting to join the corresponding IP multicast group. Upon receiving such a “JOIN” request, the network starts sending the associated multicast stream to the customer's STB.
Typically, a video stream consists of a sequence of different types of frames that includes the video and voice contents. These frames are referred to as I-frame, B-frame and P-frame. I-frame, which stands for index frame, is a frame that is used as a reference to predict and encode other frames, such as B-frames or P-frames, which follow the I-frame. An I-frame and the B-frames and P-frames that follow the I-until frame (until the next I-frame) is often called a group of pictures (GOP). The distance between I-frames in a video stream typically ranges from 0.3-5 seconds, depending on the encoding schemes and video contents. Thus, when a customer changes to a channel between I-frames, the customer may have to wait for as long as 5 seconds until the next I-frame arrives from a video server to see a clear picture. Thus, there is a need for system and method that can provide the video content faster to the customers.